Using 20s2 and hating low voltage alarm (BMS?)

[chucho]

Im using (2x) 5 4s lipo packs on my ebike but i hate each day more the voltage alarms.
One of the problems is that my voltage alarm buzzer gets energy only of two of the four cells so they become unbalanced.

So i need to be unpluging the 10 each time i stop the ebike for a while... and any way the cells get unbalanced each time i use the ebike for some time.

I have been looking for a BMS that can handle my two 20s packs but all i read is that they fail a lot and that is a big problem if you are using lipos...
I dont now realy what to do... it would be great to use a BMS so i could charge all the bateries direct and that could balance cells and cut under 3,5 and on top 4,15.
But if this is not posible i would like to find a solution for the alarms cause having cells unbalanced each time gets me very annoyed.

Thank you all for your time!
Pd: sorry for my bad english as is not my first language.

 

[hydro-one]

use conservative low voltage cutoff on controller or Cycle analyst. 3.6v per cell
use bulk charger set to conservative pack level voltage ( 4.10v per cell or less).

now use your cell checkers every day before and after you charge, outside, attended.

Alot of people do this and the cells stay in balance very well.. I personally have not balanced my 15s2p 5ah hk lipo pack in 300 or so cycles.

Do this at your own risk, charge outside ALWAYS

another sure bet would be methtek.com , if you want to sleep at night, buy methods bms with hvc breaker.

edit: I have had lipo packs burn because of neglience in checking the pack before charging. I have almost lost a family pet to a lipo fire caused by charging an over dischaged pack. You have to be sure of what you are doing if you choose the "human bms" method. But it works for me

 

[miuan]

I also use 20s lipo and have had the same issue as you report.
My solution was I stopped using the alarms altogether, and only ride with a total voltage display on my handlebar. This solution has solved all problems. I stop riding at 74V, maybe 73V if I really need to - this is unloaded voltage - and with fairly balanced pack I never see less than 3.5-3.6V on any of my cells. A brief look at the display (with throttle backed off) and you know exactly how much juice you've got left.
This is what I, and many others, use for voltage display:
http://www.ebay.com/itm/170846310527

 

[spinningmagnets]

You could also look for a DC/DC voltage regulator as a power supply, I seem to remember they were very cheap. That way you can power the alarms off of the entire pack.

 

[fechter]

It sucks they can't design these things properly. It wouldn't be that hard.

I'd love to see one that you just power from the 5v off the throttle and has near zero drain on the pack so you can just leave it always connected. My Fluke meter is like this, so I know it's possible.

The TC-54 based LVC boards or Methods' LVC boards have minimal drain and the drain is balanced. These can stay permanently connected.

Using pack level monitoring works if you're careful. It doesn't make it idiot proof, which would be the ultimate goal. Cell level monitoring would be even more important if using series-parallel pack configurations where you could bust a connection and suddenly have a cell with half capacity or worse. I've also seen cells just suddenly go bad. If you didn't pick up on the voltage drop, you could possibly reverse it or start it on fire.

 

[wesnewell]

I run 12 4s packs in a 24s2p config. I have set my LVC very conservatively to 88V so I can never go lower than 10% soc. 84V (3.5V per cell) is the ideal setting, and one of these days I'll change it to that if i ever hit LVC far enough away that I have to pedal a lot. I monitor voltage with the same volt meter as above and have never had a problem. Never saw a need for cell monitors so have never used them. I balance charge every charge to make sure the pack is in good shape. I split the pack in the middle and balance charge as 12s4p using a 14s charger I've had since running my first 14s pack. For a while I charged as 4s12p, but finally got tired of that and made my adapter cable that turned 3 4s balance plugs into 2 6s plugs. 20s using 4s packs presents problem that you can't split the pack evenly to charge it all at once. For this situation, there's a couple of different things you can do to charge the whole pack at once. One would be a bulk charger using 5 battery medics to balance and limit cell voltage on every cell. Another would be use a balancer board with a bulk charger. This would be the easiest using only one larger balance connector to the board. Cost would be about the same. Personally I'd never use a bms simply because it drains cell groups if left connected, and I've never found a lipo bms with a high enough lvc to suit me. Not to mention the extra hardware to cause more problems if/when it fails. One of these days someone may come out with a balance charger for 20-24s packs and all this will be mute. But that basically waht you'd be doing with a bulk charger and 20s balancer board.
http://bestekpower.com/balanceboard/

 

[dnmun]

why not just use a 20S lipo BMS and never worry about having to be alarmed?

if you use a BMS then you have short circuit protection on the output of the battery so when it shorts it doesn't burn up.

the BMS will also balance the pack and shut it off when you leave the bike powered up and forget it.

plus it will never overcharge if it has a BMS and you can use high current bulk chargers then too.

 

[chucho]

use conservative low voltage cutoff on controller or Cycle analyst. 3.6v per cell
use bulk charger set to conservative pack level voltage ( 4.10v per cell or less).

now use your cell checkers every day before and after you charge, outside, attended.

Alot of people do this and the cells stay in balance very well.. I personally have not balanced my 15s2p 5ah hk lipo pack in 300 or so cycles.

Do this at your own risk, charge outside ALWAYS

another sure bet would be methtek.com , if you want to sleep at night, buy methods bms with hvc breaker.

edit: I have had lipo packs burn because of neglience in checking the pack before charging. I have almost lost a family pet to a lipo fire caused by charging an over dischaged pack. You have to be sure of what you are doing if you choose the "human bms" method. But it works for me

I use i-chager 5 pack in paralel balanced charge. But methtek HVC Breaker could be an option for the future... thank you for your replay!! did not know about that product :wink:

I also use 20s lipo and have had the same issue as you report.
My solution was I stopped using the alarms altogether, and only ride with a total voltage display on my handlebar. This solution has solved all problems. I stop riding at 74V, maybe 73V if I really need to - this is unloaded voltage - and with fairly balanced pack I never see less than 3.5-3.6V on any of my cells. A brief look at the display (with throttle backed off) and you know exactly how much juice you've got left.
This is what I, and many others, use for voltage display:
http://www.ebay.com/itm/170846310527

I knew about this option but it scares me having a problem with one cell having the bateries between my legs... would not want to discover what will happen if it burst :? thanks for your replay!!

You could also look for a DC/DC voltage regulator as a power supply, I seem to remember they were very cheap. That way you can power the alarms off of the entire pack.

I dont understand this very well could you explain a little bit more? Thank you for your time

It sucks they can't design these things properly. It wouldn't be that hard.

I'd love to see one that you just power from the 5v off the throttle and has near zero drain on the pack so you can just leave it always connected. My Fluke meter is like this, so I know it's possible.

The TC-54 based LVC boards or Methods' LVC boards have minimal drain and the drain is balanced. These can stay permanently connected.

Using pack level monitoring works if you're careful. It doesn't make it idiot proof, which would be the ultimate goal. Cell level monitoring would be even more important if using series-parallel pack configurations where you could bust a connection and suddenly have a cell with half capacity or worse. I've also seen cells just suddenly go bad. If you didn't pick up on the voltage drop, you could possibly reverse it or start it on fire.

I think the same thing :wink: How hard is to do a proper BMS?!
By pack level monitoring your talking about the total Voltage indicator? im using series-parallel pack and not monitoring cells seems a bit risky for me
Thanks for your time.

I run 12 4s packs in a 24s2p config. I have set my LVC very conservatively to 88V so I can never go lower than 10% soc. 84V (3.5V per cell) is the ideal setting, and one of these days I'll change it to that if i ever hit LVC far enough away that I have to pedal a lot. I monitor voltage with the same volt meter as above and have never had a problem. Never saw a need for cell monitors so have never used them. I balance charge every charge to make sure the pack is in good shape. I split the pack in the middle and balance charge as 12s4p using a 14s charger I've had since running my first 14s pack. For a while I charged as 4s12p, but finally got tired of that and made my adapter cable that turned 3 4s balance plugs into 2 6s plugs. 20s using 4s packs presents problem that you can't split the pack evenly to charge it all at once. For this situation, there's a couple of different things you can do to charge the whole pack at once. One would be a bulk charger using 5 battery medics to balance and limit cell voltage on every cell. Another would be use a balancer board with a bulk charger. This would be the easiest using only one larger balance connector to the board. Cost would be about the same. Personally I'd never use a bms simply because it drains cell groups if left connected, and I've never found a lipo bms with a high enough lvc to suit me. Not to mention the extra hardware to cause more problems if/when it fails. One of these days someone may come out with a balance charger for 20-24s packs and all this will be mute. But that basically waht you'd be doing with a bulk charger and 20s balancer board.
http://bestekpower.com/balanceboard/

I am a bit scared of not using any indicator for the cells...
What LVC do you use? not going under a voltage would not be a problem for me... but what happens if one cell damages? everything on fire... me included i supose having cells monitored give you a plus on safety.
For charging i use one i-charger with pararel board and balancing program. Now im thinking in buying a second charger so i can charge two packs of 5 bateries.
bulk charger and 20s balancer board could be another good solution. Thank you.

why not just use a 20S lipo BMS and never worry about having to be alarmed?

if you use a BMS then you have short circuit protection on the output of the battery so when it shorts it doesn't burn up.

the BMS will also balance the pack and shut it off when you leave the bike powered up and forget it.

plus it will never overcharge if it has a BMS and you can use high current bulk chargers then too.

That would be great for me... but what BMS would you use?
I have two packs of 20s :wink:
Thank you

 

[dnmun]

i am gonna test the D166 which is built out to 20S lipo: ($55?)

http://www.bestekpower.com/74v20spcmbmspcbforli-ionli-polymerbatterypack/PCB-D166.html

and there is another one that i think is a little cheaper which also maxes out at 20S so all the channels are used:

http://www.bestekpower.com/74v20spcmbmspcbforli-ionli-polymerbatterypack/PCM-D107.html

which i think is gonna be about $25-30 when i can get enuff orders together in one group to get a good price break from them.

that's why i asked if you were in the USA. then i can forward them in bubble packs first class for cheap.

but the neat thing besides being so cheap (which is because they have a business plan of selling direct to customers instead of through the agents like BMS battery), is that they can be hacked to lower number of cells in series very easily. so if you loose a cell in your battery, you can convert the BMS from 20S to 19S, change your bulk charging voltage, then carry on like nothing happened. except for the hack. i have already hacked the D131 from 24S down to 20S to test it, then swapped it back up to 24S by replacing the resistors.

 

[chucho]

i am gonna test the D166 which is built out to 20S lipo: ($55?)

http://www.bestekpower.com/74v20spcmbmspcbforli-ionli-polymerbatterypack/PCB-D166.html

and there is another one that i think is a little cheaper which also maxes out at 20S so all the channels are used:

http://www.bestekpower.com/74v20spcmbmspcbforli-ionli-polymerbatterypack/PCM-D107.html

which i think is gonna be about $25-30 when i can get enuff orders together in one group to get a good price break from them.

that's why i asked if you were in the USA. then i can forward them in bubble packs first class for cheap.

but the neat thing besides being so cheap (which is because they have a business plan of selling direct to customers instead of through the agents like BMS battery), is that they can be hacked to lower number of cells in series very easily. so if you loose a cell in your battery, you can convert the BMS from 20S to 19S, change your bulk charging voltage, then carry on like nothing happened. except for the hack. i have already hacked the D131 from 24S down to 20S to test it, then swapped it back up to 24S by replacing the resistors.

I already knew does bms but i dont like very much the treatment for lipo bateries. Its a pity its only adjustable between 4.2V~4.35V and 2.3V~3.0V. I would like to be able to cut at 3,5v and 4,15v... that will be great for lipo life and safety. I would no care paying more if i am safer.

PD: i dont live in USA but that would not be a problem any way i have my sister living in Boston

 

[dnmun]

they are not adjustable unless you order a bunch of one tyoe i expect. they only sell one type of comparator for lipo that i saw, or tested actually.

it is set to 4.20V and will balance the cells to that voltage. i was able to pull one cell up from 4.12V to 4.20V before the BMS finally hit HVC on one channel.

this was on a new battery so it was unbalanced a lot and that was 6% of balancing in one cycle.

i balanced another 20S pack using their D131 20S lipo BMS and got it to balance to 4.20 volts too and it was made of 5200mAh and 5Ah turnigys cells. and it still balanced the pack.

 

[chucho]

they are not adjustable unless you order a bunch of one tyoe i expect. they only sell one type of comparator for lipo that i saw, or tested actually.

it is set to 4.20V and will balance the cells to that voltage. i was able to pull one cell up from 4.12V to 4.20V before the BMS finally hit HVC on one channel.

this was on a new battery so it was unbalanced a lot and that was 6% of balancing in one cycle.

i balanced another 20S pack using their D131 20S lipo BMS and got it to balance to 4.20 volts too and it was made of 5200mAh and 5Ah turnigys cells. and it still balanced the pack.

It is a pity no one does them specialy for lipo features.
Has anyone tested this bms for a long period? I am a bit scared about playing with lipos...
The balancing seems to be very good

 

[dnmun]

they are made specifically for lipo. better than balance charging in any case since they provide short circuit protection and over discharge protection too.

but group think is powerful and most cannot overcome it.

 

[chucho]

they are made specifically for lipo. better than balance charging in any case since they provide short circuit protection and over discharge protection too.

but group think is powerful and most cannot overcome it.

You are right... i was bit sleepy at that time :wink:

"but group think is powerful and most cannot overcome it" i dont understant this point

I use two packs of 20s ... do you think i could use one bms on each pack and paralel or i would have to use a conmuter or similar?

Thank you!!

Pd: The second bms is only for 30A i am using 45A controler in my ebike. The first one is of 40A... will it damage if used with my controler?

 

[dnmun]

there is the common assumption that it is best to charge the lipo cells to 4.12V to be safe and make the cells have a longer lifetime. it has just been established by constant referrals to this voltage as the correct voltage to charge to. it has not been proven, just assumed.

i think it started originally with gary and i have also assumed it was the best way to charge but now that i realize the guys who know, who build the BMSs, use the full charged state of 4.2V then i have accepted their knowledge of what works best. all the lipo laptop batteries charge to 4.20V also.

so i think the 4.20V level that these guys use to balance the BMS is appropriate and also i find the cells provide more charge if they are fully charged.

i measured the difference in stored charge between 4.12V and 4.20V to be about 6%. so if you fully charge the cells, you have 6% more capacity before it gets to the LVC, which is the risky side of the discharge, not the full charged side.

the BMS protects it on the fully charged side to prevent it from over charging.

but i consider shorts to be the biggest risk of fire and if you discharge without short circuit protection then eventually there is the risk of the battery shorting or the controller could short out and the heat generated from the short could start a fire, not from the lipo igniting itself.

i think the 40A they use for the D166 is conservative and i hope i can show that the BMS will deliver 60A if i can find a way to load up the battery that much. i think the D107 will handle 50A too.

but i want to actually test them to prove it and not assume it. that's why i am buying a buncha their BMSs to test. they are the only cheap source we have for quality BMSs now.

 

[fechter]

Let me know how that BesTech BMS works out.

One approach I've looked at several times is using a commercial unit like that and bypassing the discharge switch, which seems to always be a weak point. Instead of using the active cutoff, I'd build a simple, small interface that takes the signal from the BMS and cuts off the throttle power instead of the main power. Similarly, the small interface could cut the controller "key line" or small power wire going to the controller so even if it was left turned on for a month, the LVC would drop the load when it hit.

By bypassing the discharge FETs on the BMS, you eliminate a lot of heating and the most common failure point. You could also make it a lot smaller and less expensive.

 

[dnmun]

i also have been looking at a way to use the output mosfet signal voltage to turn off the controller but still leave the short circuit capability with the BMS output mosfets so the pack could deliver more than the amount of current allowed by the BMS but deliver it directly to the controller and bypass the BMS.

this would still allow protection from short circuits outside the controller through the overcurrent shut off of the output mosfets but the high currents needed by the controller would bypass the BMS output mosfets.

this would allow a smaller BMS to be used and still retain the balancing and the overcharge and over discharge protection. so you could use a cheap small BMS with just a few mosfets, say 15A like the D126 and the D141 i am testing.

a short inside the controller is most likely to result in an open circuit mosfet but there is still a small risk of the short in the controller wiring to the mosfets.

they do have some powerful BMSs too, that D170 model with the 40 big TO-247 mosfets must be able to handle over 250A from the looks of it. seems like they would have to build up the traces with buss wire to handle all the current it seems able to switch.

 

[chucho]

there is the common assumption that it is best to charge the lipo cells to 4.12V to be safe and make the cells have a longer lifetime. it has just been established by constant referrals to this voltage as the correct voltage to charge to. it has not been proven, just assumed.

i think it started originally with gary and i have also assumed it was the best way to charge but now that i realize the guys who know, who build the BMSs, use the full charged state of 4.2V then i have accepted their knowledge of what works best. all the lipo laptop batteries charge to 4.20V also.

so i think the 4.20V level that these guys use to balance the BMS is appropriate and also i find the cells provide more charge if they are fully charged.

i measured the difference in stored charge between 4.12V and 4.20V to be about 6%. so if you fully charge the cells, you have 6% more capacity before it gets to the LVC, which is the risky side of the discharge, not the full charged side.

the BMS protects it on the fully charged side to prevent it from over charging.

but i consider shorts to be the biggest risk of fire and if you discharge without short circuit protection then eventually there is the risk of the battery shorting or the controller could short out and the heat generated from the short could start a fire, not from the lipo igniting itself.

i think the 40A they use for the D166 is conservative and i hope i can show that the BMS will deliver 60A if i can find a way to load up the battery that much. i think the D107 will handle 50A too.

but i want to actually test them to prove it and not assume it. that's why i am buying a buncha their BMSs to test. they are the only cheap source we have for quality BMSs now.

I now that 4,2v is not the dangerous part but the experience of many people say that is better for lipo life to charge less than 4,2 the same way as charging slowly. I usualy charge to 4.15V And having it charged to 4,2V is more risk if its not very exact than to for example 4.15V

Do you think its posible to use two bms? cause i use 2 packs of 20s paralel

 

[dnmun]

yep you can use two packs in parallel under two different BMSs.

also you could tie the packs together at the cell level through the JST plugs too and then make it all one pack under one BMS by connecting the B- from each series to the B- spot on the BMS and then connect the two red output wires from the two packs together at the controller + terminal.

or you can take the two packs apart and remove the shrink wrap, solder the serial connection instead of using the big plugs and then solder the cells in parallel and attach the BMS sense wires directly to the cells at the cell level.

 

[fechter]

this would still allow protection from short circuits outside the controller through the overcurrent shut off of the output mosfets but the high currents needed by the controller would bypass the BMS output mosfets.

this would allow a smaller BMS to be used and still retain the balancing and the overcharge and over discharge protection. so you could use a cheap small BMS with just a few mosfets, say 15A like the D126 and the D141 i am testing.

a short inside the controller is most likely to result in an open circuit mosfet but there is still a small risk of the short in the controller wiring to the mosfets.

That seems like it would work. Somehow I don't trust that FET bank to reliably open in the event of a wiring or controller short. I would always recommend a fuse somewhere in series with the pack as a backup that hopefully never blows. The controller also has a current limiter that would be the primary overload protection. I have seen videos where they short the leads coming off the BMS and nothing bad happens. Little spark and it trips. I've also seen the BMS units fail a lot, but usually without fire.

 

[chucho]

yep you can use two packs in parallel under two different BMSs.

also you could tie the packs together at the cell level through the JST plugs too and then make it all one pack under one BMS by connecting the B- from each series to the B- spot on the BMS and then connect the two red output wires from the two packs together at the controller + terminal.

or you can take the two packs apart and remove the shrink wrap, solder the serial connection instead of using the big plugs and then solder the cells in parallel and attach the BMS sense wires directly to the cells at the cell level.

When are you planing to buy the bms? do you know how much battery the bms will consume when you are not using the ebike?

Ok, i understand by your explanation (sorry about my poor english ) that i could conect in pararlel each battery with another and used it like one conected to the BMS. Right? Is there any interesting post you could recomend me to read about does two option?

Thank you very much

 

[dnmun]

if you have two 20S packs, composed of 4x5S then you would connect each of the 4 packs in series and then tie the two big 20S packs together at the plus and minus terminals. but you can connect the cells in parallel using the JST plugs by making a wye with two 6pins sockets for the 6 pin JST plugs, and the open leg of the wye would connect to the sense wires to the BMS.

but!!!

the 6 pin plugs for the 5S have the negative of the first cell on the first pin so you would skip that pin when you were connecting the sense wires to the BMS to the wye you have for each sub pack.

you can see this on the 9 pin sockets i use to attach the BMS sense wires to the 7 pin 6S JST plugs

 

[chucho]

what do you think of Smart BMS?

with lcd

 

[dnmun]

what are the black and green wires over on the right side? you program it through that data cable?

 

[chucho]

I have never used it... this is what they say:

There is a bigger model for 100A but the two of them can be used for 24s
I read somewhere they were making a new version that could balance quicker